I am using 314ah Cornex grade a batteries in diy rack with a DALY bms. Can Bus connectivity is crappy, I would rather Switch to an other bms.
I have recently purchased the same Victron Multiplus-II GX together with a Pylontech US5000.
It works fine together with 8 Canadian Solar (460wp each), Enphase IQ8+ microconverters and Enphase Envoy metered-S
I installed it myself and managed to make it work off-grid too.
Regards
Bas
This may cause some consternation.
DIY bank using 18 x 280ah EVE cells - yes 18.
100% SOC is 3.45v = 62.1v.
All within the MPII 5kva, 150 and 250 MPPTâs, JK BMS and 80v fuses volt limits.
No BMS connection to the Cerbo - can, but donât want it.
Using a Smartshunt for the system SOC.
Why?
Something I learnt the expensive way.
Learnt way back after loosing my first bank to a BMS that could not do the job that I must compare the JK BMS SOC with the Smartshunt SOC.
Both must be close as damn into SOC.
When there is a large discrepancy I know there is a problem with the bank.
Have a spare 280ah cell.
Swap it with a cell in the bank every year - to use it.
Call these 19 cells my âwarrantyâ.
Can lose up to 4 cells before hitting min of 15 cells.
After 4 years of daily use all are still perfectly fine.
I have 3x 48/5000/70 Multiplus II in parallel connected to 2x 15kwh TeZe Power batteries (âŹ2200 incl tax,shipping for 30kwh)
And it works fine as the batteries simulate Pylontech.
Like others, I embarked on a DIY system a few years ago. A friend of mine had set up a group buy with a MultiPlus II 5kVA and 280Ah grade A EVE cells with an early JK BMS. I went along with that. The only thing we didnât have was an enclosure so I designed my own and had it built by a local metalworker. I bought a set of 10 units to make it worthwhile and used a few, sold a few.
I did this at a time before anyone else was making them - before China started pumping out cases to house 3.2V LFP cells.
Today, I wouldnât bother with designing your own as the economies of scale donât really work out unless you start selling, then you really need to be careful about compliance, safety certification, etc.
But decent quality LFP cells with a decent BMS thatâs properly / safely connected works wonders and will beat commercial offerings, but you need to invest some time and ensure itâs built safely.
I now have two sets of batteries: one with 16x 280Ah cells and early JK BMS, another with 16x 312Ah cells with a more recent âinverterâ JK BMS, connected in parallel. My Cerbo GX is the older one which doesnât have a dedicated CAN connection for batteries so Iâm using the serial driver and a battery aggregator which works well.
The JK BMS do have the issue with not tracking SoC well as others have mentioned so top balancing with a SoC 100% reset is required every now and then. In the sceenshot below, you can see the reported SoC is out on the first, it is more charged than 30%. Thatâll sort itself out when it next does a top-balancing process.
I have 2 diy batteries with jk bms v 19, and a SmartShunt in the DC cable, soc readings of the bms are always within 2-3% of the soc reading from the smartshunt (and always less) , I never de-charge less then 25% soc.
Verdict, you canât go wrong with either a battery from pylontech or a battery with EVE cells and a jk bms (diy or not).
I bought Winston Thundersky LiFeYPO4 batteries with yttrium 9 years ago. At present, I have 16 Ă 12 V packs, each consisting of 4 cells, connected in series and parallel to form a 24 V system, or 28.8 V nominal charging voltage. I will soon convert these batteries to a 48 V system.
The 12 V blocks were sold by Winston without a BMS. I measured them myself, and they had all been properly selected and matched with the same internal resistance. Nevertheless, I installed a top-level balancer on every cell, as well as active balancers with up to 6 A balancing current.
This type of battery has the advantage that it can handle very high currents, can still be charged and discharged at temperatures from -35 °C to 80 °C, making it suitable for winter use, and also offers a high cycle life.
Unfortunately, these cells now cost about twice as much in Europe as I paid for them at the time. The capacity is still close to 100%.
This is something most dealers and Chinese suppliers effectively do not offer today. I do not know the internal construction of Pylontech batteries, but it is conceivable that they also use this type of cell. I also remember seeing a photo showing the inside of a Victron battery. If I remember correctly, those also appeared to contain Winston cells.
The Immoveable barrier of time. Everyone claims their batteries are the best with the best cells and the best BMS, but I donât have a time machine to skip forward 10 or 15 years and see what is âstill running with good capacityâ. Pylontechs are the benchmark for longevity (if you know what you are doing when you set the system up). We are playing with several brands including TCSN and ZYC but we are so careful about our reputation. We know that Victron and Pylontech are are proven formula but we also know that we cannot become insuler to new products and technologies. The question you pose is a great one and I will follow as we all struggle with this.
How you treat your batteries matters.
Leave them in a sweltering shed at 35C+ and no battery will age well.
Undersize them or use oversized PV, pushing and pulling C rates, and it wonât pan out well over time for anyone.
Pylon have been changing a fair amount of cells lately, as have others, some sites fine, others not.
The delta is all about how they are being used.
Be kind to any battery and it will be kind to you.
You can use any LiFePo4 battery, even home made built from individual cells (see: âRay Build cool stuffâ and Andy from off-grid garage). as long as you adjust the charging parameters correctly). The discussion is weather you like to have commnunication between the battery and the Victron system (cerbo etc).
This is not stricktly necessay for operation of ESS or DESS you can get an accurate SOC from the smartshunt.
Many people are happy without communication between batt and venus.
If you connect the battery via CANbus of RS485 you can read some battery data in Venus OS (depending what the BMS is transmitting)
Victron has a list of âapprovedâ battery suppliers. Many more batteries a possible, but you have to figure out the communication protocol and the cable arrangement yourself
Already a lot of the EEL-Battery Packs with JK-BMS v15+v19 and also some of the older and newer Seplos BMS.
Many of them also 2x Packs in Parallel.
Mostly EVE MB31 A+ DoubleStud Cells some also with CATL A+ DoubleStud Cells, all 8000 Cycles.
Working like a charm! 
Pylontech Battery in EU are produce near Venice italy, you could find latest US5000 for about 760⏠+ vat on internet (etronixcenter have a very cheap delivery fee), vat for Italy solar product is 10%.
So 760âŹ/4,8= 158âŹ/KW not bad at all considering the international market price per raw Kw is still well over 120usd/Kw (not including rack electronic, cable),
Es gibt ja duzende Batterie Pack Hersteller. Ich habe mal einige angeschrieben und viele haben PACE BMS oder TDT BMS eingebaut. Hat die schon mal jemand getestet (mit Victron)?
Iâm a professional installer in France, I used to employ Pylontech US5000, but now switched to Dyness, that are equivalent in quality, but much cheaper. DL5.0C (same as US5000, but 5.12kWh instead of 4.8) is found at âŹ660.
And for bigger systems, Powerbrick plus (16kWh) is found at âŹ1800.
And one more interesting point, Dyness CAN can be switched to 250k, so if you have VEcan devices on a color control or easy solar device that has only one can port, they still can be used, works like a charm !
Would love to see photos of your setup if you are happy to share. I intend to build a similar, 3 in parallel system